Haodong Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Haodong Zhang has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Wei Ju
Sichuan University, Chengdu, China
5 shared publications - Yifan Wang
University of International Business and Economics, Beijing, China
5 shared publications - Tao Ren
Software College, Northeastern University, Shenyang, China
4 shared publications - Xiao Luo
University of Wisconsin–Madison, Madison, WI, USA
3 shared publications - Fanchun Meng
Software College, Northeastern University, Shenyang, China
2 shared publications - Siyu Yi
Sichuan University, Chengdu, China
2 shared publications
Latest Publications
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 38 (7) , 4405-4418
HGOOD-D: Hyperbolic Hierarchical Exploration for Graph Out-of-Distribution Detection
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 38 (7) , 4405-4418
HGOOD-D: Hyperbolic Hierarchical Exploration for Graph Out-of-Distribution Detection
- Open Access
- 20 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1892-1903
DisCo: Diffusion-guided Unbiased Discriminative Learning for Unsupervised Graph Domain Adaptation
This paper investigates the task of unsupervised graph domain adaptation, which facilitates the transfer of knowledge from labeled source graphs to unlabeled target graphs. Recent approaches usually utilize graph contrastive learning and pseudo-labeling to learn from unlabeled target data, which could introduce potential biased representations and supervision of target graphs resulting from serious shifts across two domains. Towards this end, we propose a novel framework named Diffusion-guided Unbiased Discriminative Learning (DisCo) for unsupervised graph domain adaptation. The core of our DisCo is to leverage both feature disentanglement and cross-domain diffusion signals to remove the potential biases for target graphs. In particular, we first utilize adversarial feature disentanglement to extract causal features that are orthogonal to domain biases. More importantly, we retrieve the labels of cross-domain source graphs to generate the conditions, which would be utilized to optimize a diffusion model for label denoising. The consistency between pseudo-labels and denoised labels is measured to reduce the potential biases during domain alignment. Extensive experiments on several real-world benchmarks demonstrate that our proposed DisCo consistently outperforms competing state-of-the-art baselines.
- 20 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1892-1903
Open AccessDisCo: Diffusion-guided Unbiased Discriminative Learning for Unsupervised Graph Domain Adaptation
This paper investigates the task of unsupervised graph domain adaptation, which facilitates the transfer of knowledge from labeled source graphs to unlabeled target graphs. Recent approaches usually utilize graph contrastive learning and pseudo-labeling to learn from unlabeled target data, which could introduce potential biased representations and supervision of target graphs resulting from serious shifts across two domains. Towards this end, we propose a novel framework named Diffusion-guided Unbiased Discriminative Learning (DisCo) for unsupervised graph domain adaptation. The core of our DisCo is to leverage both feature disentanglement and cross-domain diffusion signals to remove the potential biases for target graphs. In particular, we first utilize adversarial feature disentanglement to extract causal features that are orthogonal to domain biases. More importantly, we retrieve the labels of cross-domain source graphs to generate the conditions, which would be utilized to optimize a diffusion model for label denoising. The consistency between pseudo-labels and denoised labels is measured to reduce the potential biases during domain alignment. Extensive experiments on several real-world benchmarks demonstrate that our proposed DisCo consistently outperforms competing state-of-the-art baselines.
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1160-1170
KEGOD: Kernel-enhanced Latent Substructure Learning for Graph Out-Of-Distribution Detection
Out-of-Distribution (OOD) detection, which seeks to identify samples deviating from the In-Distribution (ID) training distribution at test time, is crucial for building robust machine learning systems. While extensive efforts have been made for Euclidean data, OOD detection on graph-structured data remains relatively underexplored. On the one hand, the specific properties of a graph may be attributed to its substructures. On the other hand, acquiring labeled data for graph learning is typically time-consuming and labor-intensive. Toward this end, in this paper, we propose a novel kernel-enhanced graph substructure learning framework termed KEGOD for unsupervised graph OOD detection. Specifically, we introduce a learnable graph generator to construct the augmented graph view that preserves distinguishable structure information. Then, for both the input graph and augmented view, a graph neural network (GNN) branch and a graph kernel (GK) branch are incorporated to explore graph latent patterns. By performing multi-branch concordance learning on the extracted graph patterns, our KEGOD captures complementary ID structural semantics in both implicit and explicit manners, enabling reliable detection of OOD graphs through semantic inconsistency. Finally, we build a self-adaptive training mechanism to automatically control diverse sensitivities of the graph patterns. Experimental results on several public graph datasets reveal the superiority of our KEGOD. Our code is available at~ https://github.com/jamesyifan/KEGOD.
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1160-1170
Open AccessKEGOD: Kernel-enhanced Latent Substructure Learning for Graph Out-Of-Distribution Detection
Out-of-Distribution (OOD) detection, which seeks to identify samples deviating from the In-Distribution (ID) training distribution at test time, is crucial for building robust machine learning systems. While extensive efforts have been made for Euclidean data, OOD detection on graph-structured data remains relatively underexplored. On the one hand, the specific properties of a graph may be attributed to its substructures. On the other hand, acquiring labeled data for graph learning is typically time-consuming and labor-intensive. Toward this end, in this paper, we propose a novel kernel-enhanced graph substructure learning framework termed KEGOD for unsupervised graph OOD detection. Specifically, we introduce a learnable graph generator to construct the augmented graph view that preserves distinguishable structure information. Then, for both the input graph and augmented view, a graph neural network (GNN) branch and a graph kernel (GK) branch are incorporated to explore graph latent patterns. By performing multi-branch concordance learning on the extracted graph patterns, our KEGOD captures complementary ID structural semantics in both implicit and explicit manners, enabling reliable detection of OOD graphs through semantic inconsistency. Finally, we build a self-adaptive training mechanism to automatically control diverse sensitivities of the graph patterns. Experimental results on several public graph datasets reveal the superiority of our KEGOD. Our code is available at~ https://github.com/jamesyifan/KEGOD.
- 1 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 36 (11) , 19644-19656
Cluster-Aware Few-Shot Molecular Property Prediction With Factor Disentanglement
Molecular property prediction plays a crucial role in drug discovery, but is always challenged by the limited number of effective labels. Compared with existing methods, we argue that the auxiliary properties of the molecule and the heterogeneous structure of different property prediction tasks have always been ignored. In this article, we propose a novel framework termed Meta-DREAM for few-shot molecular property prediction, which tailors to learning the transferable knowledge within different clusters of tasks. Specifically, we first construct a heterogeneous molecule relation graph (HMRG) with molecule-property and molecule-molecule relations to utilize many-to-many correlations between properties and molecules. The meta-learning episode can, then, be reformulated as a subgraph of HMRG. Next, we propose a disentangled graph encoder to explicitly discriminate the underlying factors of the task. In addition, we introduce a soft clustering module to group each factorized task representation into appropriate clusters and preserve knowledge generalization within a cluster and customization among clusters. In this way, each disentangled factor serves as a cluster-aware parameter gate for the task-specific meta-learner. Extensive experiments on five commonly used molecular datasets show that Meta-DREAM consistently outperforms existing state-of-the-art methods and verifies the effectiveness of each module.
- 1 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 36 (11) , 19644-19656
Cluster-Aware Few-Shot Molecular Property Prediction With Factor Disentanglement
Molecular property prediction plays a crucial role in drug discovery, but is always challenged by the limited number of effective labels. Compared with existing methods, we argue that the auxiliary properties of the molecule and the heterogeneous structure of different property prediction tasks have always been ignored. In this article, we propose a novel framework termed Meta-DREAM for few-shot molecular property prediction, which tailors to learning the transferable knowledge within different clusters of tasks. Specifically, we first construct a heterogeneous molecule relation graph (HMRG) with molecule-property and molecule-molecule relations to utilize many-to-many correlations between properties and molecules. The meta-learning episode can, then, be reformulated as a subgraph of HMRG. Next, we propose a disentangled graph encoder to explicitly discriminate the underlying factors of the task. In addition, we introduce a soft clustering module to group each factorized task representation into appropriate clusters and preserve knowledge generalization within a cluster and customization among clusters. In this way, each disentangled factor serves as a cluster-aware parameter gate for the task-specific meta-learner. Extensive experiments on five commonly used molecular datasets show that Meta-DREAM consistently outperforms existing state-of-the-art methods and verifies the effectiveness of each module.
- Top Cited Papers
- 13 February 2025
- Published by Elsevier in Information Processing & Management
- Vol. 62 (4)
MHGC: Multi-scale hard sample mining for contrastive deep graph clustering
- 13 February 2025
- Published by Elsevier in Information Processing & Management
- Vol. 62 (4)
Top Cited PapersMHGC: Multi-scale hard sample mining for contrastive deep graph clustering